In systems where high reliability is required, such as in telephony or other communication systems where a significant loss of traffic is unacceptable, it is common to provide redundant modules in the system which operate in parallel on the same inputs. Normally the output from only one such module is utilized, with the system switching to utilize outputs from the other module in the event of a failure in the original module. Since once a module fails, data outputted from that module is unreliable, it is desirable to get a failed module off-line as quickly as possible. Further, since a delay in getting the standby module on-line can result in data being lost, it is also desirable to get The standby module on-line as quickly as possible. However, there has heretofore been a relatively complicated protocol, particularly in communications applications, to communicate a failure to both modules and to then effect the necessary switching to get the failed module off-line and a good module on-line. This has resulted in either excessive switching delay, due to slow message oriented communication protocols on a shared hardware resource, or an unreasonably large number of backplane connections to control switching between the working or main and protection or standby modules. Often the number of backplane connections dictate the number of features a given module can offer. Therefore, while dedicated hardware allows for very fast switching, it also limits the modules features. However, with the high speed data transmission lines currently being utilized, this excessive delay can result in the loss of substantial data. Further, since once a module fails, it is assumed to be unreliable, any switching to a replacement module must be done without involvement of the failed module, main and standby modules should also be identical so as to be interchangeable.
It is therefore desirable that a technique be developed which substantially reduces the time delay between the detection of a module failure and the completion of a replacement module being brought on-line, while requiring a minimum number of backplane connections, and without requiring involvement of a failed module in the switching process.